Her is the question from the 70-462 Training Kit Sample Exam Questions
'You have a two-node SQL Server 2012 failover clustered instance named SYDNEY. The nodes in the cluster are named SQL-A and SQL-B. The motherboard of server SQL-B fails, and the server is offline.'
Here is what they say is the correct answer - the steps to be taken
1.Use SQL Server Setup on SQL-A to remove node SQL-B from the cluster.
2. Replace the motherboard on server SQL-B
3. Bring SQL-B online.
4. Use SQL Server Setup on SQL-A to add node SQL-B to the cluster.
According to Microsoft's documentation.
Us the following steps to recover from a reparable failure. In this case, failure is caused by Node 1 being down or offline but not irretrievably broken. This could be caused by an operating system failure, hardware failure, or failure in the SQL Server instance itself.After Node 1 fails, the FCI fails over to Node 2.
Resolve the problem with Node 1.
Ensure that all nodes are online and the WSFC service is working.
Fail over SQL Server to the recovered node.
It appears to me, that replacing a motherboard, if the operating system on the hard drives will allow such a swap and still function, is no different than rebooting a machine. All configurations, programs, settings, on the machine remain the same.
Why would you need to remove node SQL-B from the cluster? If this is true, then any time that you rebooted the SQL-B machine you would have to go thru this entire process.